The other day, I got an email from Elaine Marshall, the 2010 Democratic candidate for the U.S. Senate. Here’s the subject:
HUGE NEWS: Marshall 37% Burr 35%
The email goes on to say: “These new poll numbers come from Lake Research, one of the most respected firms in the business, and are a direct result of your hard work. It is very strong news for us – but with Senator Burr’s $6 million warchest, we still have lots more work to do.” Needless to say, the next three paragraphs asked for money.
“ORLY?”, as we say on the Internet. That phrase has a very specific, “I don’t believe a word you’re saying” connotation. Because I didn’t believe a word of this poll’s result.
I am as eager as anybody to get Senator Burr out of office. But nothing I’d heard before inclined me to think this race was anywhere near close. And certainly not in Marshall’s favor. So I went to the reliable source for political information: Intrade. There you see that Burr is trading at 80, Marshall at 19. That means, as you can read about in their “How does it work?” page, that the market believes there’s an 80% chance that Burr will be elected, a 19% chance that Marshall will be elected (and presumably a 1% chance that some un-named third party will get in).
The one rule to remember about people is that, with few exceptions, they put their money where their true beliefs are. That’s why real money prediction markets are so damn accurate: they get people to stop talking about what they want, and let them actually put their money on what they think will happen. Needless to say, that causes people to work hard to model real probabilities rather than what they want the public to think.
Now, obviously, I’m well aware that perception can become reality; if enough people start believing that Elaine Marshall is electable, then maybe they’ll start giving money to her campaign, which could actually make her electable. But that’s different than saying that a credible poll says that she’s leading the race.
I want to put my time, money, and energy in places where it will do some good. 4:1 long-shots are not my definition of “places where my money does some good”. I feel somewhat betrayed by the North Carolina Democratic party.  Are they so naive as to actually believe their 37:35 number, or do they just think that I’m gullible enough to believe it?
 But, paraphrasing W.C. Fields, at least I’m not in South Carolina